Refutation of Hans Kelsen’s concept of law

Drieu Godefridi

Introduction

Hans Kelsen (1881–1973) was the founder of legal theory in the modern era and sought to build a “pure” theory of law, in the sense of being free of political and moral considerations. In the course of an academic career that took him from Austria to Germany to the University of California at Berkeley, Kelsen left his mark on legal theory with some of its most stimulating and successful monuments. His two most ‘purely’ theoretical works are the seminal Reine Rechtslehre, published in 1934, and Allgemeine Theorie der Normen, published posthumously in 1979, which is like a series of extremely thorough and intellectually fruitful footnotes to the Rechtslehre.

By its rigour and originality, by the breadth of its field of study and its theoretical persistence, by its capacity to take into account and respond to most of the criticisms that were levelled at it — notably through polemics with that other legal theorist, Carl Schmitt — Hans Kelsen’s work structures modern thinking on law.

Kelsen claims to account for law — all possible law — as a hierarchy of norms.[1] Each norm derives its validity from the respect of norms that are hierarchically superior to it. Taken singularly, the norm is considered as an act of will, more precisely the meaning of an act of will. Thus the understanding of a particular legal system, and of any legal system understood in its pure juridicality, are possible, according to Kelsen, without involving the slightest political or moral consideration. From this point of view, legal theory is pure when it is ‘purged’ of all political and moral considerations.[2]

Two classic critiques of Kelsen’s system

Many criticisms were levelled at these attempts at theoretical purity and moral virginity,[3] starting with the obligation to describe the systems of totalitarian regimes such as Nazi Germany and the Soviet Union as legal.[4] Kelsen assumes this consequence of his vision of law.

The concept of law must account for Soviet and National Socialist law

These criticisms, which are essentially moral in nature, are fully susceptible to Kelsen’s objections and miss their objective. Reflection on the law can only be conceived of as being removed from the empire of moral philosophy. When reflection on the law proceeds from a particular conception of justice, it is no more than a derivation, in legal technique, of this conception and finds itself absorbed by moral philosophy. All too often, legal thinking ‘moralizes’ its object to exclude rights it does not like. The same tendency to moral gerrymandering of the law is found in the Christian,[5] liberal,[6] socialist and positivist traditions. What sense does it make to deny, for example, that Soviet law, which governed the behaviour of tens of millions of souls for seven decades, is law? If Soviet law is not law, since it was the only law in force on the immense Soviet territory, how can we name this set of norms that took its place? Should we consider that a country, a people, a community can persist for seven decades without law?

The problem of the Grundnorm

Another recurrent criticism of the Kelsenian system is that a legal system cannot be self-sustaining in its validity. It must by necessity be based on something other than a norm. In the sense that the ultimate norm — the tip of the normative pyramid or Grundnorm, real or fictitious — cannot, by definition, draw its validity from a norm that is superior to it.[7] This criticism, of a logical nature, is more effective than the previous one, but does not appear to be decisive, in that the whole edifice would remain valid, even if its foundation were exogenous.[8] [9]

In the latter part of the twentieth century, Profs. François Ost and Michel van de Kerchove showed that most legal systems have strange loops that deviate from the pyramid structure.[10] When a judge clarifies the interpretation of the constitution, he or she retroacts at the top of the pyramid, because his or her interpretation is binding on all actors in the legal system. However, it is not clear that these hierarchical entanglements, which are real, refute the essentially pyramidal structure of any legal system. The advantage remains, once again, with Kelsen.

The problem of generality

The flaw in the Kelsenian armour lies in its concept of generality. According to Kelsen, the norm that enshrines the arbitrariness of an authority has the highest degree of legal generality.

We begin by examining Hans Kelsen’s definition of normative generality (Chapter 1), formulating our thesis. We then examine the question of whether individual norms can be deduced from general norms (Chapter 2), which will allow us to consider one of Kelsen’s objections to the proposed concept of generality.

Chapter 1: Thesis

One must behave as prescribed by a given authority (…) represents the highest degree of generality of a norm.

Hans Kelsen, Allgemeine Theorie der Normen[11]

Although reflection on law goes back to the Greeks — Aristotle devotes some surprisingly topical reflections to it in the Politics — Hans Kelsen is credited with having revived reflection on law in the 20th century, thanks to a theory that was intended to be ‘pure’ of ideology.

The norm as the meaning of an act of will

Every norm, according to Kelsen, is either general or individual.[12] The norm is, in Kelsen’s terms, the meaning of an act of will. Norm and will are inextricably linked. The norm is not only a statement, such as a scientific proposition or a statement of fact, it is the expression of a will, more precisely the meaning — i.e. the content, which is autonomous from its sender at the moment of the statement — of an act of will. The norm is a must-be (in German, Sollen).

In one of the notes to his posthumously published General Theory of Norms, which contains the last state of his thinking, Kelsen offers a table of the different degrees of normative generality, from the individual norm in the strict sense — the most particular, concrete and immediate — to the most general norm:

“A norm is individual when the personal and material elements of the behaviour are posited as binding in concreto, i.e. as constituting the unique behaviour of a single individually determined person, e.g. the command of a father A addressed to a son B: “Close that window.” All other norms are general norms, but the general character may have different degrees:

1. The father orders his sons B, C, D: “Now go to school. Here we have a definite number of individually determined addressers in concreto, a definite number of identical unique behaviours.

2. A non-commissioned officer addresses the following command to twenty soldiers lined up in a row: “Let three men come out of the ranks! We are dealing here with a determined number of individually undetermined, thus in abstracto determined, addressers of the norm, three identical in concreto determined behaviours.

3. The Pope commands all Catholics to address a specific prayer to God on a specific day, at a specific time. This is an indeterminate but limited number of addressers of the norm determined in abstracto, an indeterminate but limited number of identical determined behaviours in concreto.

4. All men must keep the promises they have made: this is an indeterminate and unlimited number of in abstracto determined addresses, and an indeterminate and unlimited number of in abstracto determined behaviours that are identical.

5. All men must behave as Jesus prescribes. Only the authority that sets the norm is determined in concreto, the material element of the norms to be observed is not determined at all: the personal element is an indeterminate, but limited number of subjects determined in abstracto.

6. One must behave as prescribed by a given authority: only the authority that sets the norm is determined in abstracto. The material element of the norms to be followed is not determined at all; the personal element is an indeterminate and unlimited number of subjects determined in abstracto. Standard 6 represents the highest degree of the general character of a standard.”[13]

Demarcation criterion

The definition of a demarcation criterion between general and individual norms occupies legal theorists. This is not inevitable. One might consider that some norms are clearly general, others unquestionably individual, and that there is a residual category of norms which, by virtue of their abstract structure amalgamated with concrete elements, do not deserve to be described as general or individual. Such norms would be in the limbo of an in-between assumed as such.

Any norm that is not individual is general (Kelsen)

It should be noted that legal practice deals with a mass of norms whose effectiveness does not require that they be described as general or individual; the fact that they are valid norms is sufficient to ensure their effectiveness. However, we can immediately see that this option would be unsatisfactory from a reflexive point of view, in that if we propose a definition of normative generality, this definition must make it possible to create, at least by default, the category of individual norms: any norm that is not general is individual. As we have seen, Kelsen adopts this technique, reversing the terms: any norm that is not individual is general. This is formally correct and indeed allows us to consider all norms in the mode of a summa divisio.

The three criteria traditionally used by legal theorists to distinguish between general and individual norms are the generality of the addressees, the abstraction of the situation concerned and of the command as such, and the permanence of the norm.[14]

Supreme generality according to Kelsen

Thus, Kelsen considers that the fourth highest degree of generality corresponds to norms of the type “All men must keep the promises they have made”,[15] because there is an indeterminate and unlimited number of addressees defined in abstracto, and an indeterminate and unlimited number of behaviours defined in abstracto. “All men must behave as Jesus commanded” defines the fifth level of generality, before the supreme stage of the sixth level.[16]

The claim of Kelsen’s table to account for all norms, in the mode of summa divisio, its demarcation criterion (any norm that is not individual is general): none of this is, on the face of it, questionable on the formal level. It is, moreover, the whole of this picture that is indisputable, in that its internal coherence is beyond reproach and fits rationally into the reflection that unfolds, like a perfectly mastered arabesque, the General Theory of Norms.

Review

The question remains whether Kelsen’s concept of normative generality is the most appropriate one for his purpose. For the purposes of this paper, let us start with the individual norm “Close the window”, from a father A to his son B, and the fifth degree of generality, which we adapt as follows: “All children must behave as their father prescribes” (we leave out the sixth level, in order to avoid difficulties related to the status of this norm, which Kelsen himself calls a fiction):

N1: “Close the window” is an individual standard;

N2: “All children must behave as their fathers prescribe” is, according to Kelsen, a general norm.

The semantic and sender’s point of view

If we allow for the indeterminacy of the addressees, the abstraction of the situation and the command, there is no doubt that N1 and N2 occupy, indeed, two opposite positions in the field of normative generality. This is true from the semantic point of view: the individual order from A to B (N1) is radically different from N2, which is a general norm both in terms of the number and the indeterminacy of its addressees, the abstraction of the situations referred to and the indeterminacy of the commands that will follow it. This is also true from the point of view of the issuer of the norm: how can we deny that there is a world of difference, from the point of view of the issuer, between an order issued here and now, and a norm as general and indeterminate as N2, which is the mark of a legislator?

This double point of view, semantic and of the sender, is indeed the one that Kelsen claims throughout his theory. However, what about the point of view of the addressee of the norm? How is this point of view relevant? We will try to answer these two questions, which are apparently subsidiary but which open up a world of perspectives.

The point of view of the addressee of the norm

From the point of view of the addressee of the norm, there are no general binding norms, only individual norms. What is meant is that, from the point of view of the addressee of the norms, normative constraint is only actualised by individual norms, never by general norms. From the injunction of the police officer posted at the centre of a crossroads to the judgement of a court, from the point of view of effective normative constraint, the subject of law knows only individual norms.[17]

This thesis may be surprising. It postulates, first of all, that a normative order is an order of constraint — this is Kelsen’s thesis,[18] it is also ours, so we will not discuss it here.

It is only from the point of view of the normative constraint — to put it another way, of the actual application of the norm — and from the point of view of the legal subject, that there are only individual norms. “Close the window’ is an individual norm that applies, in Kelsen’s example, to son B. This individual, and strictly, character of N1 is beyond dispute. But what about N2, “All children must behave as their father prescribes”? Let us assume that N2 fits into the same normative order as N1. From the point of view of son B, what is the difference between N1 and N2? There is a big difference: the norm N1 applies to him here and now; whereas N2 does not concern him here and now.[19] From his point of view, i.e. from the point of view of the norms that effectively constrain his behaviour here and now, N2 is a non-entity, at most the vague promise of possible future individual norms, the content of which remains entirely undetermined.

In reality, N2, like any norm that is not individual in the strict sense of N1, is a collection of potential individual norms. N2 does not constrain any behaviour by itself (in the mode of constraint); in order to effectively constrain behaviour in the mode of constraint, N2 will have to be actualised in norms of type N1.

Discussion

Why take into account the point of view of the addressee of the norm? Why not limit ourselves to the dual point of view of semantics and of the issuer, as favoured by Kelsen? There is certainly a discretionary element in taking into account, or not, the point of view of the addressee. But apart from the fact that this consideration does not take anything away from the reflection, it must be stressed that the purpose of the norm is to standardise. And it is precisely from this point of view, which is the essence of the normative fact,[20] that the addressee’s point of view must be taken into account. Not to do so impoverishes, without any rational reason — that consideration which is free of any moral consideration — the reflection on the norm.[21]

If, as we argue, from the point of view of the addressee and the normative constraint, there are only individual norms, then the reflection on the concept of generality must be reopened. Indeed, from the point of view of the addressee and the normative constraint, there is no norm in Kelsen’s table, whatever its level of generality, that does not have to be actualised by individual norms (except for N1, which is already an individual norm in the strict and actual sense).

One wonders whether taking into account this point of view of the addressee does not render the question of the demarcation criterion irrelevant: if there are really only individual norms from the proposed point of view (addressee — present constraint), how can one imagine any demarcation criterion? Doesn’t this amount to distinguishing between current and deferred individual norms? All the more so since, from the addressee’s point of view, there is strictly no difference between an individual norm and another individual norm, at the moment of its actualisation. So what does this mean?

Two generalities

In truth, from the point of view proposed, the difference does not lie in the individual norm, but in the nature of the general norm from which this individual norm is derived. From the point of view of the addressee, there are, in fact, two radically distinct and irremediably heterogeneous types of general norms — we are leaving Kelsen’s picture once and for all — the general norm whose constraint content is foreseeable, and therefore avoidable, and the general norm whose constraint content is unforeseeable, and therefore inevitable.[22]

The importance of this distinction cannot be overstated; it is historically seminal to Western public law.

Psephisma and nomos

At the end of the war against Sparta, Athens, defeated, gave itself a regime that it wanted to be aristocratic and legalistic on the model of Sparta, which immediately degenerated into arbitrary tyranny. This is the regime of the thirty tyrants, who put to death, at their whim, a significant fraction of the Athenian population. The tyrants were defeated and democracy was restored. However, Athens remembers that it was under the regime of radical, and therefore arbitrary, democracy[23] that it subjugated the Aegean basin, and then declared and lost the war against Sparta. Under Euclid’s archonship, at the turn of the 5the and 4the centuries BC, reforms were implemented to curb democratic arbitrariness.

These reforms ‘constitutionalise’ existing law, placing it beyond the reach of ordinary majorities, and create mechanisms of normative control, some of which evoke, prefigure, and sometimes identify with very ‘modern’ controls of the constitutionality of laws and the legality of regulations.

The whole of this sophisticated institutional edifice is based on a distinction, now imperative, between psephisma, or decree, which is an individual norm, and nomos, or law, which is a general norm. No decree is valid that does not respect the laws in force. A decree that does not respect the laws is of no effect.

From the point of view of the addressee, generality is opposed to arbitrariness

This edifice only makes sense if one defines generality as opposed to arbitrariness, which the Athenians expressly do.[24] Through this affirmation of the necessary conformity of individual psephisma to the general nomos, Athens invented Western public law.

Just as Euclid’s reforms were intended to protect the Athenian citizen from the whims of the moment, even if they were in the majority, the whole of the so-called constitutionalist tradition is based on the requirement to remove the subject of law from the sovereign’s whim. Rule of law, Rechtsstaat, état de droit: it is a matter of demanding that power be exercised only in conformity with rules, or general norms, pre-existing its intervention, notably in the criminal field.

To which Kelsen would no doubt reply that this historical evocation is very sympathetic, perhaps true, but that his task as a theorist is not to give an account of this or that political project — which is what constitutionalism is. For him, it is a matter of developing a science of norms, which is precisely as free as possible of ideological dross.

The limits of purity in the study of law

However, this methodical rejection of the grounds of generality in the history of law shows that the desire to ‘purify’ theory has its limits, particularly in the field of the humanities (to which the theory of law and norms belongs). For it is one thing to argue that a pure theory of law must discard the constitutionalist view[25] as well as any particular ideology on law, for example Marxism. It is quite another to deny the relevance of taking into account the point of view of the addressee of the norm, on the sole ground — and one does not see any other in the course of Kelsen’s theory — that it constitutes the cornerstone of constitutionalist theory.

A concept of law that denies the perspective of the recipient is meaningless

On the contrary, we argue that no theory of law should ignore the perspective of the addressee of the norm. [26]

This amounts to asserting that Kelsen’s point of view (semantic-issue) is not false,[27] but ideological, in that it proceeds from a thirst for and illusion of purity that leads to ignoring an essential aspect of its object. One cannot treat the norm as a chemical substance or a mathematical object (i.e., by disregarding their effect on man). The norm is a human reality, whose knowledge must take into account the point of view of its addressee, because it is the object of the norm to normalise his behaviour.

Considering law without taking into account the point of view of the addressee of the norm is like considering a tool without taking into account its usefulness; it is, very strictly, a negation of its object of study.

If the point of view of the addressee of the norm is taken into account, Kelsen’s normative picture must be relegated. Indeed, Kelsen’s last two (highest) degrees of generality do not designate general norms from the point of view of their addressees — predictable and therefore avoidable — but their antithesis, i.e. collections, perfectly unpredictable and therefore unavoidable, of individual norms in potential. These norms are arbitrary, in that they merely formally consecrate the arbitrariness of an authority. [28]

Degrees in generality?

What about the idea of degrees of generality? From the point of view of the addressee, the gradation of generality seems both unnecessary and artificial: a norm — i.e., its actualisation in the form of a constraint — is either predictable or not, regardless of whether the number of addressees is limited or not, and regardless of whether the definition of its enforcement mechanism includes concrete elements. What matters is the predictability, and therefore the avoidability, of the constraint attached to it. [29]

Proposed definition of normative generality

Recognising the objective importance of the point of view of the addressee of the norm, we define as general the norm whose actualisation can be avoided by the subject of law by means of his own will. [30][31]

Whether a norm is formally individual or general — Kelsen’s picture — is a semantic consideration which, from the point of view of legal theory, is incidental.[32] Thus, the Kelsenian definition of the individual norm seems formally valid. But how can we not be surprised that a judgment is qualified as a general norm, according to the first Kelsenian degree of generality — norm 2 of his table — on the sole ground that there are at least two persons in the claim or defence[33] ? Will we say that the injunction of a police officer, posted at a crossroads, is an individual norm when it is addressed to the driver of a vehicle, but a general norm when there are two people in the vehicle? These questions show the artificiality and nominalism of the semantic debate, because everything depends on the more or less restrictive nature of the definition of individuality used. [34]

Kelsen’s purely semantic definition is not wrong, it is artificial, ‘pure’ of the essence of the reality under consideration — the norming of behaviour — ultimately very pure indeed, but this purity is that of ideology.

Role and status of generality in the Kelsene system

Now, this concept is the sun of the Kelsenian system. The concept of generality so perfectly defined by the General Theory is not defined by chance or by accident, in the course of a footnote: the Kelsenian concept of generality is of perfect coherence and continuity from the Pure Theory to the General Theory. The best illustration of this is the Grundnorm, which in its ultimate version — the last edition of the Pure and General Theory — is a fiction saying: “You must behave as the legal system commands” — which corresponds word for word to the highest degree of generality defined by the General Theory.

Kelsen’s identification of generality and arbitrariness is the conceptual substratum and structural apotheosis of his theoretical system.

Let us proceed to examine an objection in Kelsen’s theory to the above: the idea that it is, according to Kelsen, impossible to deduce an individual norm from a general norm. If this is true, then the opposition of generality and arbitrariness is meaningless.

Chapter 2. Objection

In the General Theory of Norms, Kelsen argues that it is intellectually — logically — impossible to deduce an individual norm from a general norm.[35] This surprising thesis, if true, would render the distinction between arbitrariness and generality invalid, which is based precisely on the possibility of predicting which individual norms of constraint are likely to be taken on the basis of general norms (general from the point of view of the addressee-constraint).

to be and should be

Throughout his work, Kelsen was concerned with the distinction between being and ought-to-be.[36] Rightly so, for it has been accepted since David Hume that being and ought-to-be are two distinct logical registers and that it is a mistake in reasoning to claim to deduce the norm from being; or, to put it another way, to claim to conclude from being, to ought-to-be.[37]

Thus Kelsen explains that the judgment, which is an individual norm, cannot be deduced from the law (general norm) and the facts of the case.[38] If we imagine the judge’s reasoning as a syllogism — the major of which is the law, the minor of which is the facts of the case, and the conclusion of which is the individual norm of the judgement — it appears, according to Kelsen, that it is logically impossible to deduce the individual norm from premises that certainly include a norm (the same logical register), but which also include being (a logically distinct register).

More fundamentally, and moving away from the act of judging in the strict sense, Kelsen asserts the impossibility of deducing an individual norm from a general norm,[39] independently of the facts of the case. Indeed, he explains, the individual norm is the meaning of an act of will. However, Kelsen maintains that this normative act of will, while it may be informed and inspired by various factors — such as the law — can only arise from itself, and not be derived, or deduced, from another act of will (that of the legislator, setting the general norm).[40]

Obviously, this thesis is based on a particular conception of the will. Kelsen considers that the legislator, while having willed the general norm, cannot will its application to a particular case, because “one cannot will what one does not know”[41] :

“It cannot be admitted that the legislator wants Dupuis, who caused Lelièvre’s death by a gunshot, to be imprisoned for life, because it cannot be admitted that the legislator knows that Dupuis will cause Lelièvre’s death by a gunshot. One can only want what one knows.[42]

This restrictive notion of the will is at the heart of Kelsen’s denial of the possibility of deducing the individual norm from the judgment of the general norm of the law. It must be shown to be wrong if the deducibility of the individual norm from the general norm is to be supported.

What does the legislator want?

Let us begin by noting that Kelsen is right in a sense: by definition, the legislator is not in a position to know the facts of a given case, a fortiori at the time of formulating the general norm, nor the facts of any of the cases to which his law will apply.

From this point of view, in fact, the legislature is not in a position to want Dupuis to be sentenced to life imprisonment for the murder of Lelièvre on the island of Saint-Louis on Tuesday 13 January. And it is to make up for this lack of will on the part of the legislator, which is not expressed in a particular case, that another will intervenes, that of the judge, to ‘say the law’ in the case in question.

However, Kelsen’s reasoning is problematic. No one has ever argued that the legislator was in a position to know about Lelièvre and Dupuis, or to want Dupuis to be condemned. Napoleon, who personally presided over the work that gave rise to the civilist synthesis that bears his name, is certainly not in a position to ‘want’ our contemporaries to pay themselves such and such a sum, nor to want anything from his sarcophagus in the Invalides. The mere mention of this hypothesis shows its absurdity. [43]

But whose will are we talking about? Are we condemned to the hypothesis of an omniscient legislator, in the form of Laplace’s demon,[44] so that the proposition that “the law wants Dupuis to be convicted of the murder on Tuesday in Paris, etc.” makes sense? What does the legislator want, in short? That Dupuis be convicted? Does the legislator want particular cases? Is it not obvious that the legislator’s will is concerned only with and is exhausted by the legally relevant characteristics of the situations that fall within the scope of its norm? Moreover, it would make little sense to assume that the legislature does not want its norm to apply to cases that fall within its scope.

Do we only want the present?

It is as if, for Kelsen, there is only a will in the present. Apart from the fact that this would render meaningless the very fact of positing a general norm — even in the Kelsenian sense — which, by hypothesis, concerns the future, this restrictive notion of the will entails a series of logical consequences that are difficult to sustain.

100 to Sylvie, the enforcement of this judgment is not within his competence. If Mark fails to pay, measures to enforce the judgment against him will be taken: seizure of his property by bailiff, forced sale, etc. These measures are all new individual standards. These measures are all new individual standards of constraint distinct from the individual standard of the judgment.

Now, while the judge may well have wanted Marc to pay Sylvie 100, he certainly could not have “known” or “known”, as Kelsen says, that such and such a piece of furniture belonging to Marc would be seized at such and such a place at such and such a time, sold at such and such a place at such and such a time, etc.

Does this mean that the judge not only did not want, but could not have wanted his sentence to be carried out? All this makes little sense and resembles a mind game consisting in upholding against all odds the consequences of an arbitrarily restrictive definition of the will. [45][46]

The fact remains that the general norm does not apply on its own; it can only be actualised through the intermediary of an individual norm, which is in fact the meaning of the act of another will: that of the judge. Having rehabilitated the will of the legislator, one should not deny that of the judge. The norm of the judgement is the embodiment of two wills, one deferred, because it is general and abstract — that of the legislator — the other immediate: that of the judge.

But then, how can we understand the relationship between the act of will of the judge, whose individual norm is the meaning, and the act of will of the legislator[47] , whose general norm is the meaning?

Cognitive and normative (volitional) aspects of the act of judging

By distinguishing the two aspects of the function of judging: the cognitive aspect — knowing the facts, identifying the applicable law — and the normative aspect.

On the cognitive side, which is the essential part of his function, if only from the point of view of the time and intellectual resources he devotes to it, the judge takes note of the facts of the case, translates them into the categories of the law, searches for and identifies the legally relevant standards, and finally applies these standards to the facts. Then, and only then, does the judge formulate the individual standard of the judgment: “For these reasons, sentence Marc to pay 100 euros to Sylvie”.[48]

The formulation of the individual norm, even if it embodies the act of judging, is only a subaltern aspect of the judging function. The essence of the function of judging lies in the cognitive work of the judge. And it is precisely at the cognitive level that the derivation of the individual norm takes place. The fact that the law is a general norm, a “must-be”, in no way prevents the judge from treating it, for the purposes of his reasoning, as a simple statement, an observation, a judgment of fact:

“Dupuis murdered Lelièvre” is a statement (which is not a must-be)

“The law requires that the murderer be sentenced to imprisonment” is a statement (which is not a duty to be)

“The law requires that Germaine be sentenced to imprisonment” is a statement (not a must-be)

There is no standard, strictly speaking, in this reasoning, which is a succession of observations, of statements, the unfolding of which takes place in the mind of the judge. The conclusion of this reasoning is not an individual norm in the normative sense,[49] it is not the judgement; it will only become the individual norm of the judgement when the judge, considering that his work of knowledge is complete, carries out the act of judging in the strict sense: “On these grounds, etc.”.

Kelsen distinguishes between the two aspects, cognitive and normative, of the judging function.[50] He goes so far as to concede that an individual norm can be ‘implicit’[51] and ‘contained’[52] in a general norm. But he rejects the idea that the individual norm is deduced from the general norm, or from the cognitive aspect of the function of judging, on the grounds that one cannot deduce a duty-to-be from a being, and that a duty-to-be can only be founded in itself:

“An act of thought can be linked to an act of will, and as a general rule it is insofar as if someone wants something, he must necessarily know what he wants. But this act of thought is prior to the act of will, is not identical with it nor implicit in it.”[53]

Recognition as an individual standard derived from a general standard

If the standard of the judgment is not deduced from the law, where does it come from? According to Kelsen, “to lay down the individual legal norm (…) presupposes the recognition of the general norm to be applied by the court which is competent to apply it”.[54] Recognition, which Kelsen logically locates in the will of the judge, would be the reason for the latter’s eventual application of the general norm to the singular case.[55] In other words, Kelsen distinguishes three aspects: the finding of fact and law, the recognition of the applicable law, and the judgement; the recognition being entirely within the judge’s will, and therefore the judgement, of which he would be the premise in will.

It is questionable whether this criterion, this ‘gateway’ to the recognition of the law by the court, nevertheless changes anything.

Either recognition is the result of the caprice of the court — which chose to recognise as it could have chosen not to recognise the general norm. Is it conceivable that a normative order is entirely dependent on the whim of all its courts, which is reiterated each time[56] ?

Or this recognition is itself the result of an individual norm — “we must apply the law to the case” — which the court[57] deduces, in its case, from the general norm “laws must be applied” or its equivalent (which would be enshrined, for example, in the constitution of the normative order in question). This is the very type of deduction that Kelsen contests and which, moreover, only adds a step to the reasoning that is not seen as useful.[58]

The problem is not unlike Descartes’ pineal gland. Having established the absolute heterogeneity of mind and matter, Descartes had to explain the interaction of these two realities. Pressed to clarify this point, the crucial nature of which cannot be denied, Descartes ended up formulating, in his correspondence, the hypothesis of the pineal gland, which was supposed to carry out the interaction between body and mind. However, this answer solves nothing, in that it does not explain the interaction: either the pineal gland is purely material, which explains nothing, or the interaction takes place within this gland, and everything remains to be explained.

Having shown that the recognition hypothesis merely shifts the problem without changing its nature, it remains for us to explain why the court indeed decides to apply the law, rather than its fantasy.

The Schmittian moment of the act of judging

At the end of the intellectual work of deducing the individual norm from the general norm — which Kelsen contests, and which we support — the judge could choose to formulate another individual norm, or even to abandon the intellectual work from the outset in order to apply his or her whim.

For there is indeed a decisional moment, Schmittian so to speak, at any rate Sartrean,[59] in the formulation of the individual norm; on this point, Kelsen is right, and indeed it is a truism.

Why do judges and courts apply the law — starting with their legal obligation to apply the law — instead of substituting their whim? When formulating their judgement, nothing and no one can stop them from copying a passage from the Bible or Albert Camus instead of ‘saying the law’.

This question is only one aspect of the larger question: why does the recipient of a norm respect it? Because the norm is part of an order of constraint. In other words, the addressee of a norm respects it out of fear of coercion, which takes different forms (fine, prison, reprimand, damages, loss of a right, damage to reputation, contempt, fear of reversal on appeal, disciplinary sanction, etc.). On this point, we must stop the investigation, conceding the general and psychological nature[60] of its term.

In any case, we have established that an individual norm, in the form of a statement, is deducible from a general norm, considered as a statement, and that no order of law is conceivable if this conceptual truth is not recognised.

The Kelsenian concept of generality is very consistent from the pure theory to the general theory. The deducibility of NI from NG is, on the other hand, rejected in the TG as strongly as it was affirmed in the TP. Kelsen’s evolution could not be more radical; everything he writes in the TP about the judge’s work we can take over as it stands: he clearly describes the deduction of the individual norm by the judge from the general norm (even if he does not sufficiently distinguish in the TP between the cognitive and normative aspects of the judging function). The question to be examined is not this development as such; it would be absurd to reproach Kelsen with it. The question is which of these two theses is more compatible with the Kelsenian system.

There was nothing in his definition of generality that required Kelsen to support the non-deductibility of the individual norm. Even in its highest sense, one can deduce from the Kelsenian general norm — taken as a statement — at least an individual enabling norm: according to the norm N, it is up to me, the judge, any authority, to issue an individual norm.

To conclude

“Any content can be law.”[61] Kelsen has been much criticised for this sentence from the Pure Theory of Law, because it makes a case for the ‘juridicisation’ of any legal content, be it National Socialist or Stalinist. However, this criticism, which has been repeated a thousand times[62] and is said to identify the ‘weakness’ of legal positivism, misses the point.

Firstly, because legal theory does not have to worry — on this point, our agreement with Kelsen is perfect — about the moral value of the norms it reports on. Secondly, because the weakness of Kelsen’s theory is not moral but entirely formal (conceptual) and, far from being confined to the problematic ‘point’ of the foundation in validity of the Grundnorm, it permeates his entire system.

The negation of the subject’s point of view is the source of the problem

By classifying under the term “general norms” all norms except those that are individual in the most restrictive sense defined by the General Theory of Norms — “You do that! — Kelsen shows himself to be consistent with the desire claimed as early as the Pure Theory of Law to take leave of what he calls the “subjectivist attitude” on law, consisting of thinking of law by its effects on the subjects of law. This subjectivist view, which he presents as that of the Roman jurisconsults, capitalism and lawyers, is, according to Kelsen, ideological and must be dissolved by the pure theory of law.

However, this determination to extirpate the subjectivist point of view on law leads Kelsen to define the general norm by taking into account only the point of view of the issuer of the norm, eradicating that of its addressee: any norm that is not strictly individual is general, including the norm that limits itself to consecrating the arbitrariness of an authority. Better still: the most perfectly arbitrary norm defines, according to Kelsen, the most perfectly general norm.

This paradox, which can be described more accurately as a conceptual aberration, ignores the objective difference (from the point of view of the science of norms) between a norm whose binding actualisation can be avoided by its addressee — like an electrified fence — and a norm whose binding actualisation cannot be avoided, like a cattle prod that the bearer can use as he wishes.

This objective difference in nature between two types of norms can certainly be dismissed out of hand, but in the field of a discipline, legal theory, whose very object — Kelsen dixit — is the study of norms, of normation, as an order of human conduct, its ignorance is without reason.[63]

This negation is subversive of the whole Kelsenian system

We have seen that Kelsen himself regularly used the term ‘general norm’ in a sense that semantically excludes his definition of generality.

More fundamentally, the reduction of generality to arbitrariness is subversive of the Kelsenian theoretical edifice, including the concepts of hierarchy[64] and pyramid[65] of norms.

A pyramid of norms in which each level would be limited to validating any expression of will of the lower level is easily conceivable.[66] The constitution would limit itself to establishing as law whatever the legislator designates as such, the law would do the same with the executive power, etc.

In such a way that we would have both the most general normative pyramid, and the most perfectly arbitrary.

However, this “general” purity is necessarily exclusive of any idea of hierarchy, pyramid or structure. For, if the authority finally instituted, at the base of the pyramid, can make its will — by hypothesis, any figure of its will — the norm in force, who will prevent it from revising, at its whim and at any time, the distribution of roles (levels) within the pyramid[67] ?

The identification of generality with arbitrariness is exclusive of the idea of legal order.

Knowing that the same is true of each of the floors of the pyramid, there is not and cannot be a pyramid or any order when generality is identified with arbitrariness, only an improbable tiering of pure wills that will be entwined in circles, various forms or open wars according to each person’s moods, in a mad saraband that, by hypothesis, nothing and no one will be able to transcend in law and therefore stop. The identification of generality with arbitrariness is exclusive of the idea of legal order.

If the most arbitrary norm is the most general norm, the concepts of pyramid and hierarchy of norms are meaningless — not problematic: non-existent — and law is reduced sensu stricto to the figure of a despot — in the formal, non-moral sense of living law — whose caprice permanently pulverises any idea of hierarchy. In this way, the Kelsenian ‘pyramid’, which is supposed to reflect the purest structure of law, heats up a state of nature within it.

Kelsen’s theory of judgment is a theory of pure will

Kelsen’s assertion that it is impossible to deduce an individual norm from a general norm — supported in the TGN against the TPD — also proceeds from his very particular — not false, but as amputated, and on purely ideological grounds — definition of the will.

From the fact that the author of the general norm could not have ‘willed’ that this or that particular person be condemned in concreto, and from the fact that a ought-to-be cannot be deduced from what is (true), Kelsen deduces (sic) that no individual norm can be deduced from a general norm.

The individual norm, if we follow Kelsen, is born exclusively and entirely in the obscure well of the will of the authority that enunciates it. Drawing his judgement not from the law, but from his will — exclusively from his will — the Kelsenian judge is a demiurge, an electron free of any attachment, even legal, a pure and perfect will. Kelsen’s theory of judgment is a theory of pure will.

We have shown that this theory is based on an erroneous axiomatic, because a norm can be considered as a statement (which Kelsen recognises, including in the General Theory)[68] and, as a statement, an individual norm is fully deducible — in the most elementary sense that must be conceded for the word deduce to have any meaning — from a general norm. This is the work that judges and courts have been doing since time immemorial, before carrying out the actual act of judging, which consists in transmuting the deduced statement[69] into a norm: “Not these reasons, let us judge, etc.”.

If the individual norm of judgment is always, everywhere and in its entirety — by its very nature — alien to general law, then there is no law, only a pale cloud of inert words dotted with black holes — the judges! — stirring and crushing the very idea of law.

The system of the father of modern legal positivism is not ‘problematic’ or ‘morally questionable’: it is wrong.

In a nutshell:

1. Kelsen identifies the highest degree of normative generality with arbitrariness;

2. This identification, which is not wrong in itself, does violence to the object of legal science as defined by Kelsen;

3. Kelsen constantly uses the concept of normative generality in a sense exclusive of his own definition, both in the PDT and the TGN;

4. The Kelsenian identification of generality with arbitrariness is exclusive of the notion of legal order as defined by Kelsen, in each of its components (pyramid, hierarchy): these two concepts, both structuring the Kelsenian system, cannot be reconciled;

5. The Kelsenian identification of generality with arbitrariness is exclusive of any definition of legal order that does not identify it with a pure will (an embedding of pure wills), which is the very negation of the concept of ‘order’ in its not only legal but elementary sense. Despotism is a disorder: Bertrand de Jouvenel.

Each of these propositions is structurally refutable.

Concluding hypothesis (not demonstrated by the above): does the thirst for theoretical purity, negating its object and destroying its system, which we observe throughout Kelsen’s masterly theoretical gesture, proceed from the abstractive hybris of its author, from a prior ideological vision, or both?

Selected bibliography

Andocide, “On the Mysteries”, in Discourses, trans. G. Dalmeyda, Paris, Les Belles Lettres, 2002

Aristotle, The Politics, trans. P. Pellegrin, Paris, Flammarion, 1993

Bentham, J. — Of Laws in General, London, The Athlone Press, 1970

Kelsen, H.

- Rechtsstaat und Staatsrecht”, Österreichische Rundschau, vol. XXXVI, 1913

- Théorie générale des normes, Paris, PUF, 1996

- Pure Theory of Law, Bruylant and L.G.D.J., 1999

Laplace, P.-S. — Essai philosophique sur les probabilités, Paris, Bachelier, 1814

Pauer-Studer, H. — “Kelsen’s Legal Positivism and the Challenge of Nazi Law”, European Philosophy of Science — Philosophy of Science in Europe and the Viennese Heritage, volume 17, Vienna Circle Institute Yearbook, 2014, 223–240

Plato, The Politics, trans. L. Brisson and J.-F. Pradeau, Paris, Flammarion, 2003

Tur, R. HS — “The Alternative Character of the Legal Norm: Kelsen as Defeasibilist?”, Kelsen Revisited, New Essays on the Pure Theory of Law, ed. Luis D. d’Almeida, John Gardner and Leslie Green, Oxford and Portland, Hart, 2013

Yang, K. — “The Rise of Legal Positivism in Germany: a Prelude to Nazi arbitrariness?”, Western Australian Jurist 3, Zetzel, J.E.G. (ed.), 1999, 245–257.

[1] Hans Kelsen, Théorie pure du droit (hereinafter, TPD), Bruylant and L.G.D.J., 1999 (Queen’s Rechtslehre, Vienna, 1934 and 1960), 224 ff. and Théorie générale des normes (hereinafter, TGN), Paris, PUF, 1996 (1979, posthumous), 346 ff. All references are to the French translation of the originals, published in German. (Dutch version here)

[2] TPD, 65s.

[3] In the eloquent and generic words of the American legal theorist Lon Fuller: “Jurisprudence should start with justice. I place this preference not on exhortatory grounds, but on a belief that until one has wrestled with the problem of justice one cannot truly understand the other issues of jurisprudence. Kelsen, for example, excludes justice from his studies (of practical law) because it is an ‘irrational ideal’ and therefore ‘not subject to cognition’. The whole structure of his theory derives from that exclusion. The meaning of his theory can therefore be understood only when we have subjected to critical scrutiny its keystone of negation” in “The place and uses of jurisprudence in the law school curriculum”, Journal of Legal Education, 1948–1949, 1, p. 496.

[4] Jean-René Binet, Droit et progrès scientifique, Paris, PUF, 2015, chapter 2; Herlinde Pauer-Studer, “Kelsen’s Legal Positivism and the Challenge of Nazi Law”, European Philosophy of Science — Philosophy of Science in Europe and the Viennese Heritage, volume 17, Vienna Circle Institute Yearbook, 2014, 223–240; Kenny Yang, “The Rise of Legal Positivism in Germany: a Prelude to Nazi arbitrariness?”, The Western Australian Jurist , vol 3, 2012, 245–257; Frank Haldemann, “Gustav Radbruch vs. Hans Kelsen: A Debate on Nazi Law”, Ratio Juris, vol. 18, №2, June 2005, 162–178.

[5] Christian iusnaturalism claims to discern law in nature, insofar as nature is the reflection of the divine logos. See St. Thomas Aquinas, Summa Theologiae. St. Thomas Aquinas, Summa Theologiae: Human law originally sprang from nature (91.3); human law must conform to natural law by means of reason; this natural law itself participates in what Aquinas calls the “eternal law”, a reflection of the divine logos in this world (Summa Theologiae, 91.1).

[6] Liberal iusnaturalism claims to discern right, at least fundamental rights, in the nature of man as such. See for example the two treatises on civil government by John Locke, Baruch Spinoza or Samuel von Pufendorf (De iure naturae et gentium). Although positivist in his advocacy, Thomas Hobbes maintains that man has rights as a man, already in the state of nature (Leviathan). The same ‘naturalising’ claim can be found in the contemporary form of ‘libertarianism’, in authors such as Murray Rothbard. Friedrich Hayek offers an interesting attempt to overcome this form of iusnaturalism, which he describes as naive, by arguing that if man wants to achieve certain goals, then a certain structure of law, a composite of rule of law and separation of powers, is required (see The Constitution of Liberty, 1960). Unlike most proponents of liberal iusnaturalism, Hayek assumes the ‘Weberian’ subjectivity of the values that underlie his approach.

[7] There is an extensive literature discussing the Kelsenian Grundnorm hypothesis, including Joseph Raz, “Kelsen’s Theory of the Basic Norm”, The Authority of Law: Essays on Law and Morality, Clarendon Press, 1979; J. W. Harris, “When and Why Does the Grundnorm Change? “Cambridge Law Journal, vol. 29, no. 1 (April 1971), 103–133; Graham Hughes, “Validity and the Basic Norm”, California Law Review, 59, 695, 1971; Dhananjai Shivakumar, “The Pure Theory as Ideal Type: Defending Kelsen on the Basis of Weberian Methodology”, Yale Law Journal, 105, 1996

[8] Kelsen partially takes this criticism into account in the second edition of his Pure Theory, 194 ff, by explaining that the Grundnorm is not posited but must be assumed and that therefore “the basis of its validity can no longer be the subject of a question.” (sic) This seems the very type of what Karl Popper described in his Logik der Forschung — published in the same year as the first edition of the Reine Rechtslehre, in 1934 — as an “ad hoc assumption” aimed at safeguarding an endangered theory. The same ‘assumption’ is discussed in the General Theory (references below), 343f. which aggravates its artificiality, characterising it as a fiction that ‘contradicts reality’ and is ‘contradictory in itself’.

[9] Put differently, the problem lies in the absoluteness of the Kelsenean definition of legal validity. If its concept of validity is taken from absolute to relative (intrasystemic) — no system of any kind can ever account for itself by itself — the problem disappears. How could law claim self-foundation in validity when even mathematics must concede unprovable axioms?

[10] Jalons pour une théorie critique du droit, Brussels, Publications des Facultés universitaires Saint-Louis, 1987 and De la pyramide au réseau? Pour une théorie dialectique du droit, Brussels, Facultés universitaires Saint-Louis, 2002.

[11] Op. cit.

[12] Hans Kelsen, TPD, 80 and TGN, 9. One can only agree with Kelsen when he denounces the illusion of the “traditional doctrine”, according to which law is only composed of general norms, judgments having only a declarative value (TPD, 239). The fact that the doctrine has left this illusion behind, in order to recognise that judgments are (individual) norms in their own right, is a step forward that is not unrelated to the influence of Kelsen, in particular to the analytical rigour of his distinction between being and ought-to-be. Hughes, op. cit. 695.

[13] Kelsen, TGN, 376, note 10, final emphasis added. This analysis of generality is consistent with the definition given in TPD, 245: “General norms are never more than a framework within which individual norms must be created. Only this framework can be narrower or wider. It reaches the maximum width when the positive general norm contains only the empowerment to create the individual norm, without determining its content in advance”, italics added.

[14] Jeremy Bentham, Of Laws in General, London, The Athlone Press, 1970, 76; François Ost, Sade et la loi, Paris, Odile Jacob, 2005, 274.

[15] A case in point is Article 1134 of the Belgian and French Civil Codes: ‘Legally formed agreements take the place of law for those who have made them. They may be revoked only by their mutual consent, or for the reasons which the law authorises. They must be performed in good faith.

[16] It should be noted that the sixth and highest degree of normative generality corresponds, according to Kelsen, to the fundamental norm (Grundnorm). Every legal system is necessarily based on a Grundnorm, which is a fiction (Kelsen, TGN, 339f. and 376).

[17] What about when the subject of law adapts his behaviour to the general norm, precisely in order not to be constrained? Will it be denied that, in this hypothesis, the general norm has ‘limited’ or ‘normed’ his behaviour without the intervention of an individual norm? Of an individual norm of state coercion, certainly. However, a general norm can only ever be actualised, even if in the mode of self-censorship, through the intervention of an individual norm.

[18] Kelsen, TPD, 36, 41f., 64, 220, 259; TGN, 28.

[19] N2 taken alone. If, because of N2, the son obliges himself here and now, that is, if he actualises N2 in an individual norm — “given N2, I am obliged here and now to close the window” — then he is obliged. Not by N2: by an individual norm that actualises N2 in the mode of constraint — even if it is self-generated. A general norm is only ever actualised in the mode of constraint by an individual norm.

[20] Kelsen, TGN, 115: “The object of a norm (…) is (…) human behaviour. For the norm must be observed and applied. It can therefore be adequately formulated only if it is addressed to a being who can understand its content and intends to behave in accordance with it (…)” (Kelsen’s italics, final emphasis added). This could not be better stated. See also TGN, 196, 285, 369.

[21] Taking into account the point of view of the addressee is not a choice of opportunity. The norm is a human reality through and through. To deal with the norm without taking into account the point of view of its addressee is an illusion of theoretical purity which is a negation of its object. There is only a human standard. There can therefore be no science of norms that does not scrupulously take into account the point of view of the addressee. This indicates a consubstantial link between the science of norms and anthropology, understood as a set of elementary generalisations about human nature, or more precisely about human behaviour.

[22] The difference is not one of degree, it is, let us insist, one of nature. There is an objective difference — from the point of view of the science of norms — between a norm whose binding actualisation is predictable, and therefore avoidable, and a norm whose binding actualisation is not. The former is like an electrified fence, the latter a cattle prod to be used at will.

[23] The thousand-headed tyrant described by Aristotle, in The Politics, trans. P. Pellegrin, Paris, Flammarion, 1993, 1292a.

[24] Stipulating “It shall not be lawful to establish a law for an individual if the same law does not apply to all Athenians, unless the decision is taken by 6,000 voters in a secret ballot”, see Andocides, “On the Mysteries”, in Discourses, trans. G. Dalmeyda, Paris, Les Belles Lettres, 2002, 87.

[25] Kelsen, TGN, 377.

[26] Although he implicitly takes into account the point of view of the addressee of the norm in a whole string of notions, reflections and definitions (see for example the definition of freie Rechtsfindung, or arbitrary system of law, which he opposes to the constitutionalist rule of law, TPD, 251, or the discussion of equality before the law, TPD, 146), the negation of this point of view, when discussed as such, is structuring Kelsen’s theories. Thus the polemic, in the TPD, against the notions of subjective right, legal subject and even legal person, all of which are encompassed in what Kelsen calls “the subjectivist attitude to law”, consisting of considering law “from the point of view of the interest of the parties” to determine “to what extent it harms him, i.e. threatens him with harm or disadvantage.” It is up to the pure theory of law, Kelsen explains, to dissolve this ideological viewpoint (TPD, 190).

[27] The Kelsenian definition of generality as such is not wrong, even if it leads, as we shall see, to the collapse of his conceptual system.

[28] If we accept the picture of normative generality in the TGN — which is merely the analytical deepening of the concept of generality that underlies the edifice of the two Theories — the error of the authors who establish Kelsen as the destroyer of arbitrariness — in whatever sense we take this concept (formal or material) — becomes clear. Not only does the Kelsen of the two Theories not oppose arbitrariness to law, but also the norm enshrining the purest arbitrariness of an authority embodies the quintessence (i.e., the highest degree) of normative generality. Contra: M. Cau, “Hans Kelsen et la théorie de l’État chez Dante”, trans. Pierre Girard, Droit et littérature, 5/2005. Although Kelsen had taken the view in his early writings that the rule of law was opposed to absolutism (cf. H. Kelsen, “Rechtsstaat und Staatsrecht”, Österreichische Rundschau, vol. XXXVI, 1913), but as soon as the most general norm is precisely that which enshrines the arbitrariness of an authority, this distinction is no longer possible. To write, as Emmanuel Pasquier does, that Kelsen seeks to ‘ward off arbitrariness’ is a misunderstanding: E. Pasquier, De Genève à Nuremberg: Carl Schmitt, Hans Kelsen et le Droit international, Paris, Garnier, 2012.

[29] Note, for example, that from the point of view of the addressee of the norm, there is no difference, in terms of room for manoeuvre, between the order addressed by the father to one of his sons, and the same order addressed to his three sons — the latter injunction corresponding to the first (and, according to him, weakest) degree of Kelsene generality.

[30] In any order of constraint, there are norms that are not directly constrained: this is a truism that we do not discuss here; on this point our agreement with Kelsen is perfect.

[31] The question arises as to the relationship of this antinomial generality-arbitrariness to the Kelsenian couple of general norms and individual norms. The two antinoms are difficult to combine, as their concepts do not overlap. Nevertheless, since they are two exhaustive divisions that apply to the same whole, their conjugation must be made. Among the general norms of our antinomial generality-arbitrariness, there are hardly any general norms of the fourth Kelsenian degree (such as article 1134 of the Civil Code). Among the arbitrary norms, in the sense that we have defined, we shall rank on the one hand the falsely general norms (from the point of view of the addressee), of the type of degrees 5 and 6 of generality according to Kelsen, and on the other hand the individual norms that do not conform to a general norm in the defined sense. Finally, it should be noted that an individual norm is not necessarily arbitrary, if it is ‘sourced in’ (conforms to) a general norm in the defined sense. To consider, in reverse, the intersection of antinoms in the Kelsenian framework, let us note that Kelsen’s general norms designate both general norms and arbitrary norms in the sense indicated, and that Kelsenian individual norms include, without discrimination, individual norms that are sourced in general norms (in the sense of our antinom) and individual norms that are not.

[32] Not only does Kelsen fail to recognise the primacy, from the point of view of legal theory, of normative generality over semantic generality, he does not note the existence of this duality of views (when dealing with the subject as such). In Kelsen, the semantic point of view cancels out the point of view we have called ‘normative’, which he refuses to name and distinguish, although he makes repeated implicit use of it.

[33] See in the same sense TGN, 36.

[34] From a strict semantic point of view, it is surprising that Kelsen does not at any point make use of the criterion of permanence in the definition of generality. Let us assume two identical norms; the first exhausts its validity one week after its enactment, while the second persists sine die. Will it be disputed that the second is more ‘general’ than the first? Kelsen cannot do so, because the very fact of the difference in duration means that the number of addressees of certain norms is greater (a criterion that Kelsen takes into account in his definition of the degrees of generality, but in an exclusively static mode in time, which formally there is no justification for): this shows the nominalism of a debate that is undoubtedly of interest to linguistics, but concerns law and its theory only to the margin.

[35] He had argued the opposite throughout the PDT, 82, 95, 238, 267.

[36] Kelsen, TPD, 14.

[37] Other authors are notoriously more reserved, such as Kant.

[38] Kelsen, TGN, 541.

[39] Kelsen, TGN, 542.

[40] This logically makes Kelsen what Richard HS Tur calls a “defeasibilist”, “which permits the court to disregard, disapply of defeat an applicable general norm, and (…) to substitute a new individual or general norm of its own devising (..”, in “The Alternative Character of the Legal Norm: Kelsen as Defeasibilist?”, Kelsen Revisited, New Essays on the Pure Theory of Law, ed. Luis D. d’Almeida, John Gardner and Leslie Green, Oxford and Portland, Hart, 2013, 258.

[41] Kelsen, TGN, 318.

[42] Kelsen, TGN, 335. See also TGN, 320.

[43] On this and other points, the TGN marks a conceptual radicalisation, more sophisticated than fruitful, compared to the TPD.

[44] “An intelligence which, at a given moment, would know all the forces of which nature is animated and the respective situations of the beings which compose it, if, moreover, it were sufficiently vast to submit these data to analysis, would embrace in the same formula the movements of the largest bodies in the universe and those of the lightest atom; nothing would be uncertain for it, and the future, like the past, would be present to its eyes”: Pierre-Simon Laplace, Essai philosophique sur les probabilités, Paris, Bachelier, 1814, 2.

[45] This definition of the will is no more satisfactory at the level of the general norm alone (without considering the individual norm): what sense is there in positing, as Kelsen does, that the general norms, for example of Roman law, the fruit of a thousand years of refinement by generations of prudents, praetors and then imperial officials, up to the synthesis of the Justinian Code, are the meaning of acts of will of specific persons? (see e.g. TGN, 401, note 38, which shows the ingenuity of Kelsen’s theory of the will).

[46] Although the power of reasoning and the penetrating and original character of most of its developments are to be commended, the General Theory of Norms appears, in this respect, as an analysis of particular initial definitions, which seem to proceed from a thirst (illusion) for theoretical purity that is reminiscent of both the Aristotelian circumscription of the field of possibilities and the illusions of Leibnitz’ Mathesis universalis. Cf. the way Kelsen argues in the end of the TGN that “the jurisdictional decision is not a logical inference, but (…) once it has been made and the individual norm becomes valid, it is based on the validity of the general legal norm to be applied in the concrete case. (page 567). Does one appreciate the baroque and artificiality of this retrospective foundation of the validity of an individual norm in a general norm from which it is supposedly not deduced?

[47] What is fictional, to clarify the previous development, is not the hypothesis of the legislator as such: on the one hand, this is a convenience of language that is difficult to avoid, and on the other hand, for a norm to be a norm, there must be an authority, however diverse and decentralised. The fictional case against Kelsen concerns the cognitive work of the author of general norms. When these norms are the result of centuries of refinement, it is inaccurate to present them as the result of a will, even if it is that of the legislator: this legislator is a legal fiction in the strict sense, devoid of will in the sense that Kelsen gives to the term ‘will’. The error of perspective, from this point of view, is to be compared with Kelsen’s reduction of the function of judging to the sole act of judging, cf. below.

[48] All operations perfectly described and circumscribed by Kelsen in the TPD (see note 238 and the other references cited above), before he changed his mind in the TGN.

[49] That a norm can be formulated as a statement, without normative ambition, conditions the very existence of legal theory, which gives a descriptive account of norms. See TPD, 82: “the Soll TPD, 82: “the Sollen of the juridical proposition does not have, like the Sollen of the juridical norm, a prescriptive meaning; it has only a descriptive meaning.” It is nonetheless a Sollen!

[50] Kelsen, TGN, 167s.

[51] Kelsen, TGN, 543.

[52] Kelsen, TGN, 548.

[53] Kelsen, TGN, 317. In sum, Kelsen’s argument boils down to the impossibility of deducing from the general norm to the individual norm of judgment as such. This is formally correct, at the price of a simplification of the question, i.e. of a crushing of the cognitive work by the normative moment. For if the individual norm is not normatively deducible from the general norm, it is intellectually so. This is what Kelsen recognised in TPD, 18: “a norm can be something other than the meaning of an act of will; as a significant datum it can also be the content of a pure act of thought. Some norms are willed; others are merely thought, without being willed.” To deny this concession would mean abandoning expressions such as “applying the laws” (of which Kelsen makes abundant use), which would make for a strange theory of law. See also TGN, 167, where the concession is implicit: “When a certain behaviour is objectively — i.e. in a purely cognitive way — established as conforming or not conforming to a norm presupposed to be valid (…)”, etc.

[54] Kelsen, TGN, 322, italics added.

[55] This cornerstone of the act of judging, which, according to Kelsen, is the recognition by the judge of the general norm to be applied, shows that Kelsen sometimes gives precedence to the point of view of the addressee of the norm for the purposes of his presentation.

[56] What is the difference between this system and that of natural law, in which judges are allowed to set aside a positive norm that is contrary to ‘natural justice’, and which Kelsen rightly denounced as ‘leading only to the most total anarchy’ (TGN, page 378, see also 159)? What is the difference between this system and the one that would include a norm expressly enjoining the judge to decide according to his moral-political convictions, a system that Kelsen denounced in the TPD as a “pure and simple abdication of the legislator in favour of the judge”? (TPD, 247; see also the polemic against Roscoe Pound’s system, TGN, 436, note 80). It could be argued that the recognition is even more anarchic, since no criteria are stipulated, be it of justice, be it individual, by virtue of which the judge would be led to recognise, or not, the applicable general norm.

[57] The competent court, says Kelsen, TGN, 557: one is justified in asking whether the finding of this competence is itself the result of the whim of the court seized (by chance?), or whether this court has competence by application of the general rules of judicial law.

[58] On these same grounds, Kelsen rejects the application of the rules of logic to norms more generally. For the reasons given, this rejection (absent from the PDT) seems questionable; however, the subject would take us beyond the present work.

[59] In the sense of existentialist freedom.

[60] It is arguable that it therefore falls outside the scope of legal theory in the strict sense.

[61] Kelsen, TPD, 197.

[62] Cf. the references cited above.

[63] TP, 39.

[64] Kelsen, TGN, 345s.

[65] Kelsen, TPD, 224s.

[66] Kelsen clearly asserts the possibility of this (TGN, 346), claiming to be Plato’s ideal state “which empowers judges to decide individual cases, according to their discretion, without being bound by predetermined general norms” — without being bound by legal norms, but by being bound by the consideration of the idea of the Good (a duty to be), Kelsen fails to specify (who, moreover, uses the word generality in the quoted sentence in a sense expressly irreconcilable with his definition of generality, since he excludes arbitrariness here precisely). In any case, it is worth noting that Plato himself underlines the impossibility of this ideal state, and this, let us note, by properly normative reasons (i.e., pertaining to a theory of law respectful of his anthropology): “For Socrates, how could there be someone who would be able, at any moment of life, to come and sit down with each one of us and prescribe to him precisely what it is appropriate for him to do?” (The Politics, 295a and 295b, transl. L. Brisson and J.-F. Pradeau, Paris, Flammarion, 2003). In the absence of this “god among men” (303b), we need, Plato concludes, written rules and ancestral customs, intangible and without exception (301a): where we will have recognised the concept of the antithetical generality of the Kelsenian arbitrariness.

[67] One would be tempted to answer by distinguishing, as Kelsen does, between norms that empower to create lower norms (e.g. the constitution empowers the legislature to create laws), and norms that prejudge the content of the lower norms to be created (e.g. excluding the race criterion, or limits to freedom of expression: these examples are from the Kelsen of TGN). The latter could be perfectly arbitrary/general, in the Kelsenian sense, by not prejudging the content of the lower norms at all, whereas the former are not (it is the legislator who creates the laws and he alone, in strict compliance with the conditions laid down by the constitution). However, this distinction is not conclusive in this case. Either the normative order under consideration provides for the possibility of its own revision, a power that ultimately falls to the lower normative level in the considered scheme of “perfect” generality/arbitrariness. Or it is not perfectly general/arbitrary in the Kelsenian sense.

[68] The fact that a norm can be regarded as a statement is a condition for the possibility of any theory of law, which Kelsen already emphasises in Pure Theory, 82 ff, when he distinguishes between the norm as a ‘prescriptive Sollen’ and the statement about the norm, which is a ‘descriptive Sollen’. (Strictly speaking, the expression ‘descriptive Sollen’ is a contradiction in terms: a normation is not a description but a normation; one would therefore more readily speak of ‘description of a Sollen’ than of ‘descriptive Sollen’).

[69] In the sense of grasping the statement in and through an act of will whose meaning is judgment.

Firts published here